An evaluation of the use of the dry-weight-rank and the comparative yield biomass estimation methods in paramo ecosystem research

6 Pages·1993·1 MB·English

by  Hofstede, Robert G M

Caldasia 17(2); 205 -210, 1993
AN EVALUATION OF THE USE OF THE
DRY-WEIGHT-RANK AND THE COMPARATIVE YIELD
BIOMASS ESTIMATION METHODS IN PARAMO
ECOSYSTEM RESEARCH
ROBERT G.M. HOFSTEDE
HENK J.Lo.,WITTE o,
Department Systematics, Evolution and Paleobiology, Hugo de Vries-Iaboratory, University Amsterdam,
Kruislaan318, 1098SMAmsterdam, TheNetherlands.
Resumen
Se evaluó la aplicabilidad de una combinación de dos métodos para estimar la biomasa
general y la composición botánica, en una vegetación natural paramuna en el Parque
Nacional Natural los Nevados (Cordillera Central de Colombia). El primer método (ecom-
parative yield») determina labiomasa general, destruyendo parcialmente lavegetación de los
cuadrantes demuestreo yelsegundo (<<dryweight rank») determina lacomposición botánica
con base enelpeso seco, sindestruir lavegetación. Estos métodos, inicialmente desarrollados
para pajonales forrajeros en Australia, se adaptaron para ser utilizados en el ecosistema
paramuno. Como resultado se obtuvo una estimación de labiomasa aérea de 2864 9peso
seco. m·2(desviación stándard 48) en lacual, lagramínea Calamagrostis effusa contribuyó
con el70%.
Puede concluirse que elmétodo de producción comparativa es útil para estimar labiomasa
en el ecosistema paramuno, siempre ycuando se utilicen las adaptaciones mencionadas.
Por otra parte laestimación de lacomposición botánica con este método dio resultados más
satisfactorios que con elmétodo «dry weight rank», elcual presenta demasiados problemas
debido alacomplejidad de los pajonales del páramo.
Abstraet
The use of the combination of the semi-destructive comparative yield method for overall
biomass estimation and the non- destructive dry-weight-rank method for studying botanical
composition on adry weight basis inan undisturbed páramo vegetation inthe Los Nevados
national park (Colombian Central Cordillera) was evaluated. These methods, developed for
Australian production grasslands, were adapted for use in the páramo ecosystem. The
average above ground biomass inthe area was estimated as2864 9dryweight. m-2 (sd.48),
ofwhich the bunchgrass Calamagrostis effusa contributed with ca 70%.
When used with some adaptations, thecomparative yield method seems suitable for biomass
estimations inthe páramo ecosystem. The here presented estimation ofbotanical eomposi-
tion with this method gave better results than dry-weight-rank method, which had too many
shortcomings for use inthe complex páramo grassland ecosystem.
Introduction
crease if the same confidence interval is to be
The most widely used method to estimate maintained (Wiegert, 1962).
aboye ground biomass in grasslands is the
destructive sampling of plots and subsequent The vegetation structure of neotropical equa-
extrapolation to establish the yield of vegeta- torial alpine grasslands (páramos) is one of
tion quadrats (Cochran 1963; 'TMannetje, the most complex among the grasslands in the
1978). The optimum quadrat size is that size world (Cleef, 1981). The dominant growth
that provides the smallest confidence interval forms of the páramos in the Colombian Cen-
ofthe mean for agiven labour cost. When the tral Cordillera are stemrosettes (Espeletia
vegetation structure becomes more complex, hartwegiana subsp centroandina), constitut-
the optimum quadrat size and cost will in- ing an emergent (up to 5 m) vegetation layer
205
CaldasiaVol. 17,1993
aboye high (up to 1m) growing bunchgrasses Both the selection of the standard quadrats as
(Calamagrostis spp., Festuca spp), and dwarf the assignment of scores is based on visual
shrubs (Pernettya spp., Bacharis spp.). An parameters as density, height and woodiness
additional lower vegetation layer may be of the vegetation. An initial training period is
present with ground rosettes (Hypochaeris necessary to calibrate visual selection and
spp.) sedges (Carex spp.) and short-growing scoring with actual dry weight and to mini-
grass species (Calamagrostis spp., Agrostis mise differences in scoring by different inves-
spp.). Almost bare spots are common tigators (testing of scores by harvesting and
(Salamanca, 1991). Due to this spatial hetero- weighing, Friedel, et al., 1988).
geneity and complex structure, optimum quad-
rat size for destructive sampling will be high. Botanical composition on a dry weight basis
For accurate biomass measurements ofpáramo can be estimated with the non-destructive
vegetation, a less destructive and faster dry-weight-rank method ('TMannetje & Ha-
method, which takes into account the spatial ydock, 1963). In this method, alarge number
heterogeneity, may be favourable. In the of quadrats are selected at random in an area.
present study we evaluate the usefulness of a In each an estimate ismade ofthe species with
combination of two methods to estimate bio- highest rank on a dry weight basis, by means
mas s and composition of Australian produc- of the mentioned visual parameters. Simi-
tion grasslands, adapted for and applied in a larly the species with the second and third
relatively undisturbed páramo vegetation. rank are identified. For each species the load
of first ranks is multiplied by 8.04, second
The comparative yield/dry weight rank ranks by 2.41 and third ranks by 1.00 (sensu
method 'TMannetje & Haydock, 1963). Summing the
resulting products gives aspecies-score, which
In the semi-destructive comparative yield is then expressed as a percentage of summed
method for estimating total aboye ground species-score of all species. This percentage
grassland biomass (Haydock & Shaw, 1975), represents the botanical composition. "TMan-
a series of sampling units (standard quadrats) netje & Haydock (1963) have shown that the
is selected subjectively and marked to con- mentioned multipliers can be applied to any
struct an ordinal scale over the total range of type of grassland.
dry matter yields within a study area, using
the following procedure: First, a sampling The two methods, inc1uding the multipliers,
unit with the highest dry matter yield of the have been applied in grasslands all over the
area under investigation (class 5) and a sam- world, and have been tested and improved by
pling unit with the lowest yield (not zero, different authors (Friedel et al., 1988; Kelly
class 1) are selected. Subsequently, a sam- & McNeill, 1980).
pling unit in between 1 and 5 (class 3) and
sampling units in between 3 and 5 (class 4) Adaptations to the paramo ecosystem
and in between 1and 3 (c1ass 2) are selected.
In this study the comparative yield/dry-
A large number of sampling units (scored weight-rank combination was adapted to esti-
quadrats) is selected at random in the area, of mate overall aboye ground biomass (inc1uding
which dry matter yield is given a score on the litter, adult stemrosettes were excluded) and
scale of the standard quadrats. Depending on botanical composition of an undisturbed
the relative difference between the scale units, páramo vegetation (Caño de Agua Leche, Los
intermediate scores are allowed. Finally the Nevados National Park, Colombian Central
five standard quadrats are harvested and the Cordillera at 4100 m alt., 4 48'N, 75 24'W).
material oven dried. With the biomass of In an area of 3ha with arelative homogeneous
these five quadrats and the scores of the ran- distributed vegetation five standard quadrats
domly selected quadrats total biomass can be (l m2) were se1ected and marked. Next the
estimated (Haydock & Shaw, 1975). whole area was divided in a systematic
206
Hofstede &Witte: Yield biomass
transeet-line pattern. While walking along all 50
transeets, 50 quadrats were randomly seleeted L (0.91 * Sei +4.68)
by throwing (blind-eyes) a hoop (diam.100 i=1
cm). By this proeedure all points in the area In B=-------- (1)
had the same ehange to be sampled. Eaeh 50
quadrat was given a seore on the standard
quadrat-seale, with a preeision of 0.25 (1.00,
1.25, 1.50, 1.75,...5.00). Although 50 quad- where: B= Average aboye ground biomass of
rats should suffiee to minimise eonfidenee area (g DW.m-2)
limits (Friedel et al., 1988), two independent Sei = Seore of seored quadrat ion range 1-5
data-sets of 50 were analysed to obtain a The average aboye ground biomass of the
better idea about the stability of the results. data-sets 1and 2was 2836 and 2891 gDW.m-2
respeetively (average of both sets: 2864 g
Average seores of the eomparati ve yield DW.m-2, sd=48).
method ofthe 2 data-sets were 3.28 and 3.18,
with standard deviations of 0.88 and 1.04 The estimates of the biomass presented here
(range: 1-4.75), refleeting the heterogeneity were eompared with the yield oftwo indepen-
of the vegetation. The seores approximate to dently seleeted and harvested seored quadrats
a normal distribution (p<0.05). (seores: 2.75 and 3.25). With the regression
equation their biomass was estimated as 1316
After harvesting and drying (24 hr at 105 C), and 2074 g respeetively. Aetually they sup-
the yield of the 5 standard quadrats was ported 1228 and 1966 g, an over estimation of
weighed. Total yields (dead and living, inel. 7%.
litter) amounted to 234, 808, 1788, 4275 and
9462 gDW.m-2. After alogarithmie (In) trans- Dry-weight-rank estimates were also made in
formation of these values, the funetion InY = the previous seleeted seored quadrats. Ranks
4.68 + 0.91 * Qe deseribed the relation be- were assigned to 5 species and to 4 groups
tween standard quadrat class (Qc)and Inyield (dwarf shrubs, ground rosettes, sedges and
(In Y), (R2 = 0.99, p<O,OI). This funetion short grasses were not ranked on species level).
eould be applied to the s ores of the seores Calamagrostis effusa eontributed 70% to to-
quadrats to ealculate average dry weight.m-2 tal biomass of the area with data-set 1 and
of the area: 74% with data-set 2 (table 1). For the other
Table 1. Load of first, second and third ranks per species, species scores and botanical composition (%of dry
weight) ofthe overall biomass, asobtained bythe dry-weight-rank method. Results of2data-sets of 50 scored
quadrats.
Data set 1 Data set 2
* *
Species/groups 1st #2nd 3rd Specles % dry # 1st #2nd #3rd Specles % dry
ranks ranks ranks seore welght ranks ranks ranks score welght
Calamagrostis effusa 42 24 6 401.5 70.1 42 33 8 425.2 74.3
C. recta 5 13 6 77.5 13.5 5 O 1 41.2 7.2
Sedges 1 4 8 25.7 4.5 O 5 8 20.1 3.5
Short grass 1 2 4 16.9 2.9 O O 2 2.0 0.4
Ground rosettes O 1 4 6.4 1.1 O 2 10 14.8 2.6
Rosaceae 1 3 10 25.3 4.4 3 2 7 35.9 6.3
Lupinus microphyllus O O 1 1.0 0.2 O 2 4 8.8 1.5
Dwarf shrubs O 2 4 8.8 1.5 O 5 6 18.1 3.2
Juvenile Espeletia O 1 7 9.4 1.6 O 1 4 6.4 1.1
Total score 572.5 100.0 572.5 100.0
207
Caldasia Vol. 17, 1993
species differences between the two data-sets other methods can not be made. Neverthe-
were considerable. less, some remarks can be made.
Another botanical composition estimation The comparative yield method seems suitable
procedure was developed to be used with the for páramo ecosystem research. Because of
comparative yield method. After harvesting the big difference between the yields of dif-
the standard quadrats the yield was separated ferent spots within the vegetation it is easy
in the different species/groups, which were even for relative inexperienced observers to
then subdivided in living and dead material. select the 5 standard quadrats covering the
The contribution of a particular unit (e.g., biomass range, and to score vegetation quad-
Calamagrostis effusa, young leafs), expressed rats on this scale. This is suggested by the
as apercentage of the total yield of astandard small diference between the two sets and by
quadrat was determined. Subsequently, bo- the accuracy of the estimated yield of the two
tanical composition was ca1culated as: sampled scored quadrats. Ain-transformation
of the yields allows the use of a linear regres-
5 sion model to describe the relation between
2,
yields and scores, but one yield per c1ass is too
(%Sc(a+x)*(%Vt(a)*(l-x) +%Vt(a+l)*x)}
few for statistical analysis. More replicas per
a=I
class or more classes, possibly with a smaller
%Vt -----------(2)
quadrat size, would improve the significance.
100
The variance of data-set scores was quite
high, which reflects the inherent heterogene-
where %Vt = Contribution (%) of unit to the
ity of the vegetation and its biomass. Analys-
overall biomass in the area
ing the distribution of the scores with respect
= to additional parameters (terrain factors, land
%Sc(a+x) Contribution (%) of score a+x to
use) recorded during the scoring of the quad-
total number of scored quadrats
= rats could probably provide useful informa-
a Integer part of score (l,2,3,4,5)
= = tion about the distribution of the biomass over
x Fractional part of score (if a 4, x 0.25,
0.5,0.75; else x = O) the area.
=
%Vt(a) Contribution (%) of unit to the yield
Although the composrtion of the standard
of standard c1ass a.
= quadrats was very diverse, the estimates of
%Vt(+1) Contribution (%) of unit to the
overall botanical composition with the com-
yield of standard quadrat c1ass a+1
parative yield method resulted in detailed fig-
ures. Results of this kind, or even more
This method resulted in amore detailed over-
precise, can be obtained using conventional
all botanical composition than the one ob-
methods, but that involves the destruction of
tained by the dry-weight-rank method: tomore
far more quadrats and a considerably larger
species acontribution was attained and a sub-
time investment.
division per unit was made in living and dead
standing crop (table 2).
The dry-weight-rank method for estimating
botanical composition seems less appropriate
Discusion and conclusions
in this vegetation, with its complex structure,
especially as information on the living/dead
Rigourous statistic testing was not possible ratio s can not be obtained with this method.
due to of the small sample size, the lack of Due to the dominance of Calamagrostis effu-
standard quadrat replicas and the application sa in almost al! scored quadrats, the contri bu-
of the methods injust one páramo site. More- tion of the other species to overall biomass
over, data on biomass from other regions are was negligible. Of the 150 rank scores made
not available so comparisons of the present in each data-ser, 75 pertained to Calamagros-
overall biomass figures with the results of tis effusa and 75 to all 8other species/groups.
208
Hefstede &Witte:Yieldbiemass
Table 2. Botanical composition (%ofdryweight) ofthe five standard quadrats and ofthe overall biomass (results
of 2data- sets of 50 scored quadrats), as obtained by the comparative yield method.
Botanlcal composltlon orstandard quadrats (%) Overoll botanleal
composltlon (%)
Unit
Standard Standard Standard Standard Standard Data-set Date-ser
Quadrat 1 Quadrat 2 Quadrat3 Quadrat 4 Quadrat 5 2
Living
Calamagrostis effusayoung leaves* 1.0 1.6 2.6 2.5 2.2 2.3 2.2
C.effusa old leaves ** 1.3 4.5 4.5 4.7 3.8 3.5
C.effusa sheath 5.0 12.4 8.1 7.9
C.effusa shoots 0.8 0.5 0.1 0.1
C.recta young leaves * 0.4 23.2 0.1 0.1
C.rectaold leaves ** 0.5 11.5 0.2 0.2
C.recta sheaths 1.8 11.7 0.7 0.5
Sedges 8.5 4.1 2.7 2.8
Short grass 1.1 0.1 0.1
Ground rosettes 7.9 6.2 1.1 0.2 1.7 2.2
Rosaceae 3.0 0.5 2.8 1.3 1.2
Lupinus microphyllus 5.6 0.1 0.8 1.1
Other herbs1.00.50.30.3
Dwarf shrubs 32.4 13.7 1.1 0.3 4.1 5.7
Juvenile Espeletia 2.2 0.2 0.2 0.3 0.3
Dead
Calamagrostis effusaleaves 3.7 15.8 14.1 n.o 12.1
C.effusa sheaths 10.6 12.8 4.8 4.7
C.rectaleaves 1.0 0.4 0.3
Sedges 17.4 1.7 2.9 3.8
Short grass 2.6 0.1 0.2
Juvenile Espeletia 1.0 0.1 0.1 0.1
Litter
Calamagrostis spp .32.7 41.8 42.7 54.4 33.6 45.6 44.4
Other 10.7 3.6 14.7 0.2 6.7 6.1
Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0
* Overall green leaves are defined as young leaves.
** Green leaves with yellow spots are defined as old leaves.
This is reflected in a low difference of the All other species showed different contribu-
proportions of Calamagrostis effusa and high tions in the results of the two methods.
differences of the proportions of the other
species between the two data-sets (Table 1). The comparative yield method, with the men-
Both the results of the botanical composition tioned adaptations, could be a fast and reli-
obtained in the comparative yield method able method of estimating biomass and
(summation ofliving and dead, litter exc1uded) composition of complex grasslands, if tested
and the results obtained inthe dry-weight-rank more intensively. The execution ofthe present
method, show percentages of Calamagrostis study was ca. 24 hours, inc1uding harvesting,
effusa in the same range (CY:67% -set 1-and subdividing and weighing, which may be con-
62% -set 2-; DWR 70 and 74% respectively). sidered fast to obtain this kind of information
209
Caldasia Vol. 17.1993
(Wiegert, 1962). When the sample set is Dissertaciones botanicae 16.J. Cramer,
increased asmentioned (more classes ormore Vaduz.
COCHRANW.G. 1963. Sampling Techniques. John
replicas per c1ass) nonparametric regression
Wiley &SonsoNew York.
and bootstrap analysis (Efron & Tibshirani, EFRONB,.&R.TIBsHIRANI1.991. Statistical analysis
1991) could be a recommendable test. An inthe computer age. Sclencie253:390-395.
important additional reason for using it in FRIEDEL,M. H., V.H. CHEEWINGS& G.N. BASTIN.
1988. The use of comparative yield and
threatened systems is that the methods re-
dry-weight rank techniques for monitoring
quires less quadrats to be destructed than grassland. J.Range Manage. 41: 430-435.
conventional methods. The dry-weight-rank HAYDOCKK,.P.&N.H.SHAW.1975. Thecomparative
method seems tohave toomany shortcomings yield method forestimating drymatter yield of
to be practical in the páramo ecosystem. pastures. Aust. J.Exp. Agr. Anim. Husb. 15:
663-670.
KELLY,R.O. & L. McNEILL. 1980. Tests of two
Acknowledgements methods for determining herbaceous yield
and botanical composition. Proc. Grassl.
The authors would like to express their thanks to Soc.Afr.15: 167-171.
the Instituto Nacional de los Recursos Naturales 'TMANNETJEL,. 1978. Measurement of grassland
vegetation and animal production. Com-
Renovables y del Ambiente (Inderena; Bogotá
monw. Agricult. Bureaux. Farnham Royal,
Manizales) for providing hospitality in the Los
Bucks.
Nevados National Park. Dr. Alejandro Velázquez
___ • & K.P. HAYDOCK.1963. The dry-
Montes and G. Guillot M. Se. are acknowledged weight-rank method forthebotanical analysis
for their constructive criticisms. Dr. Sonia of pasture. J. Brit. Grassl. Soco18:268-275.
Salamanca improved the spanish abstract. S. SALAMANCAV. 1991. The vegetation of the
páramo and its dynamics in the volcanic
The first author was supported by grant W84-283 massif Ruiz-Tolima (Cordillera Central, Co-
ofthe Netherlands Foundation fortheAdvancement lombia) Ph. D. Dissertation, University of
of Tropical Research (WOTRO). Amsterdam.
WIEGERTR,.G. 1962. The selection ofanoptimum
quadrat size for sampling the standing crop
ofgrasses and forbs. Ecology43: 125-129.
Literature cited
CLEEF,A.M. 1981. The vegetation of the pára-
mos of the Colombian Cordillera Oriental.
210